Perceptual Vector
Quantization of Binary Images for Content Based Image Retrieval
Rishi Mukhopadhyay and Ishwar K. Sethi
Intelligent Information Engineering Laboratory
This summer’s research was an extension of the work done in
conjunction with Aiyesha Ma last year through the Uncore program (last
year). This year’s work will be
extending that work to demonstrate the viability of using the shape descriptors
we developed in pattern recognition tasks and more specifically in
content-based image retrieval.
Additionally, I will have to work further with last year’s methods to
refine the technique, increase robustness to noise and address reviewer
comments.
Last year’s work was accepted for presentation at the
International Knowledge Engineering conference in Las Vegas, June 23rd-26th.
Rishi Mukhopadhyay just completed his second year of studies in Computer Science at the Engineering school in Cornell University. His non-academic interests include film, reading, driving aimlessly and other explorations of cyclic futility, graduating on time, and long moonlit walks on the beach.